Beszel: The Lightweight Monitoring Stack Your Homelab Actually Needs

I used to run Prometheus, Grafana, and a half-dozen exporters just to see if my containers were healthy. It worked. But it also consumed 2GB of RAM, required constant query tuning, and felt like using a sledgehammer to check a pulse.

Then I found Beszel.

Beszel is a lightweight, modern monitoring stack built specifically for the "I just want to know if things are working" crowd. After running it alongside my Prometheus setup for a month, I made the switch permanent. Here's why it might be the monitoring tool your homelab has been waiting for.

What Is Beszel?

Beszel is an open-source monitoring solution that combines a lightweight agent with a clean web dashboard. It's designed for small-to-medium deployments — think homelabs, small VPS fleets, or a few Docker hosts — where full observability platforms feel like overkill.

The pitch:

• Single binary for the agent (or one Docker container)
• Pre-built dashboard — no query writing required
• Automatic service discovery for Docker containers
• Alerts via Discord, Slack, Telegram, or webhook
• Minimal resource usage (~50MB RAM vs Prometheus's 1GB+)

The Problem with Prometheus in Homelabs

Don't get me wrong — Prometheus and Grafana are powerful. They're also powerful enough to be frustrating:

For a homelab with 5-10 services, Prometheus often creates more work than it saves. You're maintaining a monitoring system and the things it's supposed to monitor.

Setting Up Beszel

Beszel runs as a hub-and-spoke model. The hub is the web dashboard and data store. The agents run on each host you want to monitor.

Step 1: Deploy the Hub

The hub stores metrics and serves the web UI. SQLite is the default database — no external dependencies needed.

# docker-compose.yml
services:
  beszel-hub:
    image: henrygd/beszel-hub:latest
    container_name: beszel-hub
    ports:
      - "8090:8090"
    volumes:
      - ./beszel-data:/data
    environment:
      - BESZEL_PORT=8090
      - BESZEL_DB=/data/beszel.db
    restart: unless-stopped

Deploy and visit http://your-server:8090. Create your admin account on first launch.

Step 2: Add Agents

On each host you want to monitor, run the Beszel agent:

# docker-compose.yml on monitored host
services:
  beszel-agent:
    image: henrygd/beszel-agent:latest
    container_name: beszel-agent
    environment:
      - BESZEL_KEY=your-public-key-from-hub
      - BESZEL_HUB=http://your-hub-ip:8090
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock:ro
      - /proc:/proc:ro
      - /sys:/sys:ro
    restart: unless-stopped

The BESZEL_KEY comes from the hub dashboard — click "Add System" and copy the key.

Note: The Docker socket mount enables automatic container discovery. Remove it if you only want host-level metrics.

Step 3: Proxmox LXC Setup

For Proxmox LXCs, you have two options:

Option A: Privileged LXC with nesting

# On Proxmox host
pct set <vmid> -features nesting=1

Then deploy the Docker agent normally.

Option B: Binary install (no Docker)

# Inside the LXC
wget https://github.com/henrygd/beszel/releases/latest/download/beszel-agent-linux-amd64.tar.gz
tar -xzf beszel-agent-linux-amd64.tar.gz
sudo mv beszel-agent /usr/local/bin/

# Create systemd service
sudo nano /etc/systemd/system/beszel-agent.service

[Unit]
Description=Beszel Agent
After=network.target

[Service]
ExecStart=/usr/local/bin/beszel-agent -key your-key -hub http://your-hub:8090
Restart=always

[Install]
WantedBy=multi-user.target

sudo systemctl enable --now beszel-agent

What You Get Out of the Box

Once agents connect, the dashboard shows:

• CPU usage (per-core and aggregate)
• Memory utilization with breakdown (used, cached, buffers)
• Disk usage per mount point
• Network I/O with bandwidth graphs
• Docker containers — status, CPU, memory per container
• System temperature (if sensors available)

All with 1-minute resolution and 30-day retention by default.

Configuring Alerts

Beszel's alerting is refreshingly simple. In the dashboard:

1. Go to Settings → Notifications
2. Add your Discord webhook, Slack URL, or Telegram bot token
3. Set thresholds per system:
   • CPU > 80% for 5 minutes
   • Memory > 90%
   • Disk > 85%
   • Container down

No routing trees. No silence configurations. Just "tell me when something breaks."

Example Discord notification:

🚨 Alert: plex-server
CPU usage: 94% (threshold: 80%)
Duration: 6 minutes

Real-World Migration: My Setup

I migrated a 5-host homelab from Prometheus to Beszel over a weekend:

Before:

• Prometheus + Alertmanager + Grafana: 3 containers, ~1.5GB RAM
• Node exporter on 5 hosts
• cAdvisor for Docker metrics
• Custom dashboards I never quite finished

After:

• Beszel hub: 1 container, ~80MB RAM
• Beszel agent: 1 container per host, ~30MB each
• Working dashboards immediately
• Alerts that actually made sense

Time to full migration: 3 hours including testing.

When to Stick with Prometheus

Beszel isn't a wholesale replacement for every use case. Keep Prometheus if:

• You need long-term metric retention (years, not weeks)
• You're doing complex aggregations across hundreds of hosts
• You need custom business metrics (application-level, not just infrastructure)
• Your team already knows PromQL and has built dashboards

For everyone else — especially homelab operators with 2-20 hosts — Beszel removes complexity you probably don't need.

Advanced: Running Beszel Behind a Reverse Proxy

If you want HTTPS and a custom domain:

Nginx Proxy Manager:

• Forward beszel.yourdomain.com to http://beszel-hub:8090
• Enable Websocket support (required for live updates)

Traefik:

labels:
  - "traefik.enable=true"
  - "traefik.http.routers.beszel.rule=Host(`beszel.yourdomain.com`)"
  - "traefik.http.routers.beszel.tls.certresolver=letsencrypt"
  - "traefik.http.services.beszel.loadbalancer.server.port=8090"
  - "traefik.http.routers.beszel.middlewares=auth@file"  # optional basic auth

Caddy:

beszel.yourdomain.com {
    reverse_proxy localhost:8090
}

Conclusion

Monitoring shouldn't be a second job. Beszel proves you can have visibility into your homelab without running a full observability platform.

If you're currently maintaining a Prometheus stack just to check if Plex is up, give Beszel a try. The 30 minutes you spend setting it up will save you hours of dashboard tuning and query debugging.

Resources:

• Beszel GitHub
• Official Documentation
• Docker Hub

Running Beszel in your homelab? Found a clever integration? Drop a note in the Discord — always curious how people are monitoring their setups.